The eating disorders (EDs) anorexia nervosa (AN) and bulimia nervosa (BN) are severe psychiatric disorders with high mortality. While AN is mainly characterized by severe underweight and BN individuals are at normal to high weight and regularly binge and purge, there are many overlapping symptoms across both disorders, such as food restriction, episodic binge eating, purging or excessive exercise. Little is known about the neurobiology of EDs and how specific ED behaviors are linked to particular biological mechanisms. The NIMH Funding opportunity Advancing Eating Disorders Research through Dimensional Studies of Biology and Behavior is in direct response to this lack of knowledge. In the present application, using the Prediction Error construct from the NIMH's Research Domain Criteria (RDoC) project, we examine how brain function and structure could be biological mechanisms underlying dimensions of ED relevant behaviors. This construct has been associated with specific brain reward circuit responses, which could be related to particular ED relevant behaviors. Aim 1. is to study how functional neural reward circuits underlie dimensions of ED pathology using the RDoC construct prediction error. This application will study all individuals ages 16- 29 years who present for ED treatment to capture the full range of ED pathology (conforming with RDoC requirements), to study the age range with the highest mortality rate for AN, and to identify potential developmental factors by studying adolescents and young adults while staying within a limited age range. The study will include detailed clinical assessments as well as functional magnetic resonance brain imaging (fMRI) for taste-reward prediction error response. ED relevant behaviors will be predicted by anteroventral striatum (caudate/putamen) as well as insula prediction error response. In addition, lower body mass index (BMI) in ED subjects will be associated with higher activation to reward value predicting conditioned stimuli in the anterior cingulate cortex, an area important for comparing options of behavior response to stimuli including reward. Aim 2. is to perform structural magnetic resonance brain imaging (sMRI) for gray and white matter volume, as well as diffusion tensor imaging for white matter integrity. We expect that more severe ED behaviors will predict lower caudate and insula volumes. In addition, gray matter volume in insula and ventral striatum will predict strength of functional brain reward activation from Aim 1. Aim 3. is to cluster ED individuals based on prediction error response instead of traditional DSM criteria, and compare ED relevant behaviors based on this new classification with age matched healthy adolescent and adult controls. Prediction error response will be superior to DSM in clustering of ED subjects by dimensions of sensitivity to salient stimuli and in contrast to controls. This will suggest that EDs are disorders of altered salience response.